一种融合乌鸦搜索算法的K-means聚类算法
A Hybrid K-means Clustering Algorithm Integrating Crow Search Algorithm
高海宾1
作者信息
- 1. 淮南联合大学信息工程学院,安徽淮南,232001
- 折叠
摘要
传统的K-均值聚类算法(K-means)对初始聚类中心的选择敏感,容易陷入局部最优解,并且需要预先设定聚类数量K,这在实际操作中往往难以实现.为了解决这些问题,提出了一种融合乌鸦搜索算法的K-means聚类算法.该算法利用乌鸦搜索算法的全局搜索能力,自动确定最佳的聚类数目K,从而提高聚类的质量和效率.通过在Seeds数据集进行实验计算卡林斯基-哈拉巴斯(Calinski-Harabasz)指数等评价指标,发现该算法聚类效果明显优于传统的K-means算法.
Abstract
The traditional K-means algorithm is sensitive to the selection of initial cluster centers,prone to falling into local optimal solutions,and requires the pre-setting of the number of clusters K,which is often difficult to achieve in practical applications.To overcome these issues,a K-means clustering algorithm that integrates the Crow Search Algorithm is proposed,aiming to address the limitations of the traditional K-means algorithm in clustering analysis.This algorithm leverages the global search capability of the Crow Search Algorithm to automatically determine the optimal number of clusters K,thereby improving the quality and efficiency of clustering.Experiments conducted on the Seeds dataset,calculating evaluation indices such as the Calinski-Harabasz index,have found that the clustering effect of this algorithm is significantly superior to that of the traditional K-means algorithm.
关键词
K-means算法/乌鸦搜索算法/聚类/Calinski-Harabasz指数Key words
K-means algorithm/crow search algorithm/clustering/Calinski-Harabasz index引用本文复制引用
基金项目
安徽省教育厅自然科学研究重点项目(KJ2021A1313)
安徽省教育厅自然科学研究重点项目(2023AH051161)
安徽省高等学校省级质量工程项目(2022zygzsj047)
出版年
2024
NETL
NSTL
维普
万方数据